Code with OpenCode and DeepSeek instead of Claude Code
A developer switched their coding agent from Claude Code to OpenCode with DeepSeek, reporting comparable quality at far lower cost for coding sessions.
Done withClaude Code
- What they did
- The poster switched their coding agent from Claude Code to OpenCode, running DeepSeek as the model. The post is a short Bluesky note and gives no setup detail. It doesn't describe configuration, tasks or review steps. It says they made the switch the day before and then coded with it.
- How it went
- They report quality comparable to Claude Code at a much lower price, about 25 centavos of real for 2 hours of straight coding work. No benchmarks or task details are given.
- Worth knowing
- This is one user's impression after about a day of use, so check quality on your own projects before relying on the cost comparison.
Try it yourself with Claude Code
Help me try a cheaper coding setup. Walk me through installing OpenCode and connecting it to DeepSeek with my API key, then run it on a small real task in [my project folder], such as [a bug fix or small feature]. When done, summarize how the output compares with what I get from Claude Code and roughly what the session cost.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: Bluesky · Oct 14, 2026
Related use cases
Turn tickets into merged pull requests
Pulls tasks from GitHub Issues or Linear, then runs Claude Code or Codex in isolated Kubernetes pods to carry each one through to a PR.
Reverse engineer Claude Desktop to find a path to Linux support
A maintainer pointed Claude at the minified Claude Desktop app to find why Cowork only runs on macOS; it found the VM and sandbox layers and mapped a route to Linux.
Automate event social posting across platforms
Claude Code connected social channels to drive event registrations, working well on Twitter and Reddit and somewhat on LinkedIn.
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Investigate broken builds and merge fixes
OpenAI's internal OpenClaw agent Androidclaw uses Git, GitHub and logs to investigate broken builds, trace problems to incidents, and sometimes prepare and merge fixes.
Run agents that work Jira tickets, review PRs and send updates
Luke runs several Hermes agents that read Jira tickets, write code, review PRs, track costs, summarize docs and send updates.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.